How do I monitor opencode sessions?

Install ClawMetry (pip install clawmetry). The daemon auto-detects the opencode SQLite store at ~/.local/share/opencode and renders every session, tool call, model, and token in one dashboard at http://localhost:8900.

Does ClawMetry catch opencode's parallel-provider double-charge?

Yes. When opencode fans a turn to Anthropic + OpenAI + local, each provider bills separately; ClawMetry parses the per-provider usage records and shows the fan-out cost on the Cost tab, plus flags a session where the total provider bill is far larger than the returned tokens suggest.

Is ClawMetry safe against a running opencode?

Yes. The SQLite database is opened with mode=ro (read-only) via SQLite URI mode; if the database is locked by opencode's writer we retry once after 100ms. No write lock is ever taken.

Does ClawMetry work with opencode's TUI and CLI modes?

Yes. Both write to the same SQLite store, so ClawMetry picks them up identically. Sessions are labelled with their surface (TUI or CLI).

How much does ClawMetry cost for opencode?

Free forever on your own machine. Cloud starts at $9/node/mo (Starter) or $19/node/mo (Pro).

Observability and cost tracking for opencode

See every opencode session, tool call, and token as it happens. ClawMetry reads the opencode SQLite store under ~/.local/share/opencode and renders sessions, tokens, cost, and per-provider routing in one dashboard. No SDK.

Supported runtime Reads opencode SQLite session store Read-only Free trial

What ClawMetry sees on opencode

ClawMetry's opencode adapter opens the store opencode already writes (~/.local/share/opencode) in read-only mode and surfaces:

Sessions: started at, last activity, model, message count, per-session token totals.
Live activity: every tool call as it lands, with args and result. Turn-level tokens split.
Per-turn tokens: Per-turn tokens computed by opencode: input, output, cache-read, cache-write, reasoning.
Cost per session, per model, per hour: cache-aware where the model supports caching.
Multi-provider parallel routing (opencode fans a hard turn to multiple providers): the opencode-specific signal ClawMetry surfaces that generic dashboards do not.
Approval and kill switch: watch for token-maxing signals (message repeat, no-progress heuristic, cost-per-hour threshold), pause or kill before the next call.

Source: the adapter is at clawmetry_pro/adapters/opencode.py. Every parse rule and every SQL SELECT is in that file.

Enable it in about two minutes

1. Install ClawMetry

Pick the line that matches your OS. Windows users: paste the install.cmd line, not the shell one.

# macOS or Linux
curl -fsSL https://clawmetry.com/install.sh | bash

# Windows (cmd.exe or PowerShell)
curl -fsSL https://clawmetry.com/install.cmd -o install.cmd && install.cmd

2. Run the daemon

clawmetry

The daemon auto-detects opencode's local store and starts reading immediately.

3. Open the dashboard

Open http://localhost:8900. Switch the runtime chip to opencode.

Start the free trial →

The one thing this catches that opencode alone does not

Left alone, opencode's failure mode is a fanned turn charged three times because every provider took the request. ClawMetry watches for it: cost per hour on the opencode runtime, message-repeat detection, and per-session token-growth curves. Set a threshold, and the approval queue pauses before the next call. Kill switch (Pro) stops it.

All 19 runtimes ClawMetry observes

opencode is one of 19. Free on every plan: OpenClaw, NVIDIA NemoClaw. Paid on Cloud Pro or Self-Hosted Pro: the rest.

OpenClaw NVIDIA NemoClaw NanoClaw PicoClaw Claude Code Codex Cursor Aider Goose opencode (this page) Qwen Code Hermes Pi Deep Agents n8n Antigravity GitHub Copilot Grok QM